All free indices below are spatial, and . Contract the vacuum Ricci tensor using :
Reversing both antisymmetric curvature pairs identifies the last term as . The spatial Gauss–Codazzi equations for a spatial hypersurface then give
For the magnetic part of the Weyl tensor, moving the normal to the first slot of the metric volume tensor introduces a minus sign:
The remaining curvature indices are spatial, so the normal projection in the Gauss–Codazzi equations for a spatial hypersurface gives
The interchange makes the two spatial covariant derivative terms equal. Thus the constants, with the paper's orientation and extrinsic-curvature convention, are
In particular, the sign of the magnetic expression must include the minus from moving past the first volume-tensor index. These formulas reconstruct the spatial electric part of the Weyl tensor and magnetic part of the Weyl tensor from the induced metric and the extrinsic curvature of a spatial hypersurface.
The electric part of the Weyl tensor has one normal in each antisymmetric pair of the Weyl tensor. Therefore
For the magnetic part of the Weyl tensor, contracting the first slot gives by antisymmetry of the metric volume tensor; contracting the second gives by antisymmetry of the last curvature pair. Consequently both tensors are entirely spatial:
These spatial projection tensor identities actually hold for any Weyl tensor; the vacuum assumption is needed subsequently to replace it by the Riemann curvature tensor.
For the Vacuum Einstein equations with zero cosmological constant, , and its tensor contraction gives . Every Ricci tensor and scalar curvature term in the Weyl tensor consequently disappears:
Thus the Weyl tensor contains all vacuum curvature. Here vacuum means Ricci-flat; a nonzero cosmological constant would retain a constant-curvature contribution to the Riemann curvature tensor.
Use the usual normalization that is mean-zero unit-variance white noise orthogonal to the past, and extend periodically in . A nondegenerate causal periodic autoregressive model of order one requires . Write , with season indices understood modulo .
For , multiply the recursion by and use orthogonality of the current innovation to the past. The periodic Yule-Walker equations are
At lag zero, expansion of the squared recursion gives
In particular, , so an equivalent variance equation is . Negative lags are obtained from the given symmetry .
Replacing covariances by their sample versions gives
Take the nonnegative scale and require a positive estimated predecessor variance. Using matched seasonal sample covariance matrices makes the quantity under the square root nonnegative by Cauchy-Schwarz inequality; with unmatched covariance estimates one can instead fit seasonal ordinary least squares and use the nonnegative residual variance. Statistical consistency of the covariance estimates gives statistical consistency of these plug-in estimators. If the driving noise variance were not normalized to , the equations would identify , not the two scales separately.
First, the absolute values in the printed limit are an error: the correct long-run variance of a stationary process is the signed sum of autocovariances. Directly,
so
Each coefficient tends to and has absolute value at most . Absolute summability and the dominated convergence theorem therefore give
The sum of absolute values is an upper bound, not the general limit. For a concrete counterexample to the printed assertion, take with unit-variance iid noise. Then , , and other covariances vanish. The absolute sum is , but , so .
For correlated data the central limit theorem can still hold under suitable strong mixing of a stationary process and moment conditions, but the limiting variance is the long-run variance of a stationary process, rather than the one-observation variance. When it is positive,
Positive serial dependence usually increases the standard error, while negative dependence can reduce it. The approximate effective sample size of a stationary sample is when that denominator is positive. A long-memory time series can require a different normalization or a different limit law. Absolute covariance summability alone does not prove a CLT: if with a common independent random scale taking values and with equal probabilities and iid with the standard normal distribution, the off-diagonal covariances are zero, yet has the nonnormal scale-mixture law . The example is not an ergodic stationary process. If the long-run variance of a stationary process is zero, the usual nondegenerate square-root- CLT is unavailable.
For the causal AR(1), and
The conditional distribution follows because the current innovation is independent of the past.
For a fixed known , condition on the observed and use the transitions . Their conditional maximum likelihood criterion is, up to constants,
Differentiating in yields
Since ,
Thus it is an unbiased estimator, even conditionally on , and
It has statistical consistency with mean-square convergence, and the iid-noise strong law of large numbers also gives almost-sure statistical consistency. Its conditional distribution is exactly normal with the displayed mean and variance. If is also observed and all transitions are used, replace by .
The fixed- qualification is necessary for the exact finite-sample claims. If is jointly estimated, conditional likelihood function is linear regression with an intercept: writing and , the unconstrained estimators are
This ratio is not generally an unbiased estimator and does not have the preceding finite-sample variance. Under the usual stationary regression conditions it has statistical consistency; its asymptotic variance is . Indeed, it differs from by , an asymptotically negligible endpoint term. Profiling an unknown innovation variance does not change the fixed- estimate of .
The printed strong white noise assumption does not imply a normal distribution or even a finite fourth moment. Thus it does not, by itself, determine the covariance of a quadratic transform. We give the intended Gaussian calculation, and then the general finite-fourth-moment answer.
If is Gaussian, is a centered Gaussian process. Put and . Since and , the centered transform is
The supplied Hermite polynomial identity makes the cross terms vanish and gives
Equivalently,
The constant has no effect on covariance.
For a general iid noise with , put and , its fourth cumulant. Independence and expansion of third and fourth moments give, for ,
where
Summing the geometric series explicitly gives
The covariance of quadratic transforms of a linear process follows from a fourth-moment expansion consisting of the three Isserlis theorem pairings, plus the fourth-cumulant contribution when all four noise indices coincide. The third-moment contribution similarly requires three coincident indices. This proves the general formula without assuming a normal distribution. For Gaussian white noise , recovering the simpler answer. If and the noise has infinite fourth moment, need not have finite variance, so an autocovariance function may not exist.
Use the original-noise coefficients and for . The autocovariance function of a causal linear process is , since the original noise has variance . Thus
For ,
By covariance symmetry,
In particular, all nonzero-lag covariances are negative, despite the positive autoregressive coefficient.
Expanding the stable autoregressive inverse gives the infinite moving-average representation in linear innovations:
Therefore
For comparison, the causal representation in the originally supplied noise is
Both converge in , since their coefficients are square summable. The second is causal, but its driving noise is not the linear innovation process.
The root-reflected, invertible autoregressive moving-average model is
This follows by multiplying the filter identity for by . Its moving-average root and autoregressive root are both outside the unit disk. Thus it is the representation in terms of the innovation process, rather than merely another noise representation with the same spectrum.
Here the innovation process consists of linear innovations, the linear one-step prediction errors: , where is the closed linear span of the past 's in . For Gaussian processes this is also the conditional expectation prediction error. For general non-Gaussian strong white noise these two notions can differ.
The given is not the linear innovation process. The moving-average factor has its zero at , inside the unit disk, and is noninvertible as a causal moving-average filter. The identity
is the moving-average root reflection that places this zero outside the unit disk. Thus the causal invertible representation has innovation variance , rather than the given variance .
For an explicit verification, define
The filter has constant squared modulus , so is weak white noise with variance . The new moving-average factor has root and is invertible, while its autoregressive factor is causal. Hence the past spans of and agree, and belongs to that past span. Orthogonality of to past therefore identifies it as the linear innovation process. The variance difference proves that it cannot be . If the original noise is Gaussian, the new linear innovations are independent Gaussian variables; without Gaussianity they need only be uncorrelated random variables.
For the spectral questions take the unique weakly stationary process solving the equation, as is customary for a stable autoregressive moving-average model. The equation alone also admits nonstationary solutions differing by , for which a spectral density of a stationary process need not exist. Let be the backshift operator. Since the autoregressive root is , the stationary solution is causal and
Using the convention , the spectral density of a stationary process is
The stationary mean is zero, since .
Apply the tower property of conditional expectation to the intensity equation. Since is a specified individual hazard function,
As in the conditional equation, this describes the first-order expected increment of the observed counting process.
With independent censoring, the at-risk process is . Independence and the common censoring survivor function give
Without censoring, , so its unconditional expectation is survival:
The distinction between and is immaterial for a continuous event time.
In the counting-process intensity in survival analysis, the counting process jumps once if an observed event occurs; the at-risk process is one immediately before only while individual is still observed and event-free. Conditional on that information, the chance of an event in a short interval is for an at-risk individual and zero otherwise. This is the meaning of
The history is the filtration generated by the observed event and censoring histories, entry information, and available covariates strictly before . In particular, it determines the current risk set but does not reveal future event times. The right-hand side is the infinitesimal compensator of a counting process; the equality is interpreted as an intensity statement to first order in .
An individual has neither died nor been censored by time exactly when and . Under independent censoring, this has probability . Thus the probability of leaving observation by either route is
The two pieces agree at . The expression counts the first of death and censoring, so it does not double-count individuals who would eventually experience both.
For administrative censoring with uniform entry, let entry time . Then has the uniform distribution on , where , , and . Its density and survivor function are
Dividing density by survival and integrating gives the censoring hazard and integrated hazard:
The hazard function is undefined once no uncensored individuals remain. As , it diverges like and the cumulative hazard function diverges logarithmically. This reflects the certainty of administrative censoring by the longest possible follow-up time, not dependence of censoring on the event process.
The three independent sample means satisfy . A linear transformation of their joint multivariate normal distribution is again normal. The two treatment contrasts have covariance
Writing and , their joint distribution is
The positive correlation coefficient arises from the shared control arm. Under positive variances, .
The braid group has generators and relations
In the strict monoidal category , put and send
The images are invertible because the Yang–Baxter operator is invertible. Distant images commute by the tensor interchange law, and adjacent images satisfy the braid group relations by the Yang–Baxter operator equation tensored with the remaining identities. Thus these assignments define group homomorphisms , and hence a functor .
Juxtaposition of braids puts generators into the corresponding two blocks of tensor factors, so their images are the tensor products of their images. Strictness of gives and exactly. The resulting functor is strict monoidal and sends to .
For an object of a braided monoidal category, take its self-braiding
It is invertible by definition. Suppress canonical associators using the monoidal coherence theorem, and put , . The hexagon identity gives
Apply naturality of this braiding to the morphism in its second argument. It says
Substitution gives
This is the equation for a Yang–Baxter operator. Restoring the uniquely determined associators gives the non-strict diagram in the paper. Every object therefore has the canonical Yang–Baxter operator supplied by its self-braiding.

Pinned article: Introduction to the OurBigBook Project

Welcome to the OurBigBook Project! Our goal is to create the perfect publishing platform for STEM subjects, and get university-level students to write the best free STEM tutorials ever.
Everyone is welcome to create an account and play with the site: ourbigbook.com/go/register. We belive that students themselves can write amazing tutorials, but teachers are welcome too. You can write about anything you want, it doesn't have to be STEM or even educational. Silly test content is very welcome and you won't be penalized in any way. Just keep it legal!
We have two killer features:
  1. topics: topics group articles by different users with the same title, e.g. here is the topic for the "Fundamental Theorem of Calculus" ourbigbook.com/go/topic/fundamental-theorem-of-calculus
    Articles of different users are sorted by upvote within each article page. This feature is a bit like:
    • a Wikipedia where each user can have their own version of each article
    • a Q&A website like Stack Overflow, where multiple people can give their views on a given topic, and the best ones are sorted by upvote. Except you don't need to wait for someone to ask first, and any topic goes, no matter how narrow or broad
    This feature makes it possible for readers to find better explanations of any topic created by other writers. And it allows writers to create an explanation in a place that readers might actually find it.
    Figure 1.
    Screenshot of the "Derivative" topic page
    . View it live at: ourbigbook.com/go/topic/derivative
  2. local editing: you can store all your personal knowledge base content locally in a plaintext markup format that can be edited locally and published either:
    This way you can be sure that even if OurBigBook.com were to go down one day (which we have no plans to do as it is quite cheap to host!), your content will still be perfectly readable as a static site.
    Figure 5. . You can also edit articles on the Web editor without installing anything locally.
    Video 3.
    Edit locally and publish demo
    . Source. This shows editing OurBigBook Markup and publishing it using the Visual Studio Code extension.
  3. https://raw.githubusercontent.com/ourbigbook/ourbigbook-media/master/feature/x/hilbert-space-arrow.png
  4. Infinitely deep tables of contents:
    Figure 6.
    Dynamic article tree with infinitely deep table of contents
    .
    Descendant pages can also show up as toplevel e.g.: ourbigbook.com/cirosantilli/chordate-subclade
All our software is open source and hosted at: github.com/ourbigbook/ourbigbook
Further documentation can be found at: docs.ourbigbook.com
Feel free to reach our to us for any help or suggestions: docs.ourbigbook.com/#contact